Circle U. King’s College London Seed Fund Project - Understanding students’ use of speech recognition and translation technologies in their learning

Project Lead: Dr Karl Nightingale, Programme Director, BSc. Biomedical Science  King’s College London

Project Partners: Univ. Prof. Dragoş Ciobanu, Professor of Computational Terminology & Machine Translation,  Centre for Translation Studies, University of Vienna, Alina Secară, MA PhD, Senior Scientist, Centre for Translation Studies, University of Vienna, and Janeesh Hans, MA, Project Assistant, Centre for Translation Studies, University of Vienna

Project duration: February - September, 2023

Project funding: Circle U. King’s College London Seed Funding and UniVie internal funding

Project aim and methods: The project aimed to characterise second language learners’ perceptions, motivations and use of a range of technologies in their learning, ranging from lecture capture and transcription to automated machine translation software. This was examined in two contexts:

(i) International students in UG programmes in a range of disciplinary contexts (King’s College), [and]

(ii) Second language learners (typically German-speaking students belonging to a variety of nationalities learning in German and English) in post-graduate and undergraduate programmes (UniVie). 

We used one-to-one interviews and/or focus groups to understand the advantages and challenges that current technologies/workflows used in (i) lecture captioning / transcription (e.g. automated speech recognition +/- human editing) and (ii) automatic speech translation offer students, including:

- the extent and factors that influence students’ use of lecture captions and/or transcripts;

- the extent and factors that influence students’ use of automatic speech-to-speech translation and/or text-based machine translation;

- any barriers to their effective use (e.g. accuracy, speed, formatting on devices, faculty practice etc.).

Activities: A mixture of face-to-face and online interviews and focus groups were carried out with over 50 undergraduate and postgraduate students from a range of disciplinary / national backgrounds (students from Mainland China, Hong Kong, Middle East, and Continental Europe). Students from a wide range of disciplines were interviewed, and understanding the differences in disciplinary-specific practices enriched the study findings (e.g. King’s contrasting students’ uses in BSc Computer Science vs. BNurs Nursing programmes). The Vienna strand of the study started with Translation Studies students, and then expanded to include Law, Social Science, and Computer Science students, as well.   

Outcomes: A thematic analysis of the interview transcripts will inform a report and/or a joint research paper. In particular, we have generated understanding / insight into the extent and ‘drivers’ that encourage the students’ uptake and/or use of the surveyed technologies, the variation in the students’ critical engagement with machine-generated material, and the students’ views towards their lecturers’ use and tolerance of surveyed technologies in teaching, as well as assessment. 

The unanticipated emergence of AI technologies in 2023 resulted in students frequently mentioning ChatGPT as the most popular example of Generative AI when discussing machine translation / academic English supportive technologies. We felt this was an appropriate extension to the project, and have gathered insights into students’ use of a variety of AI technologies in their study. We would now like to expand on this and we are therefore planning an application for ‘follow on’ Circle U. Seed Funding Scheme in October 2023. This would also include a third perspective from a further Circle U. partner, namely UC Louvain who confirmed interest. 

Impact: This project maps onto several Circle U. initiatives, notably the ‘Shaping the Future of Higher Education in a Changing World’ White Paper which seeks to encourage pedagogic innovation. The project findings are likely to give insight into some of our students’ practices within the Circle U. network, and represent a ‘snapshot’ of student / staff perceptions and needs at a critical time of the emergence of AI technologies in HE. Insight from this project can help inform policy on the use of lecture capture, transcription, and automated machine translation at Institutional, Faculty, and Programme-specific levels within the Circle U. network. Locally, the results will also be shared with the UniVie Centre for Teaching and Learning and the UniVie Digitalization Working Group.